Abstract

The growth of cities is closely linked to the overall economic growth of nations. Especially in urban planning, predicting and modelling the growth trajectory of cities is crucial for ensuring sustainable economic growth. The growth of cities brings with it many social and economic gains, but it also increases many economic and social demands. The growth of cities is analysed based on geographical foundations, the assumption of increasing returns, and the random growth approach. The growth of cities is usually discussed in terms of three different models. The random growth approach is called Gibrat's law, and this approach allows cities to be analysed empirically. Gibrat’s law is analysed comparatively with parametric analyses as well as nonparametric and semiparametric models. This study aims to examine the validity of Gibrat's law at the provincial level in Türkiye using parametric and semiparametric panel data models. The study utilizes annual data from 2007 to 2019 at the provincial level. The analysis reveals that parametric models provide weak evidence for the validity of Gibrat's law, while semiparametric models provide stronger evidence.